D365 F&O at 90% - What I Wish I Knew at 10%

We're 90% through our Dynamics 365 Finance & Operations migration at Julius Silvert. The light at the end of the tunnel is real this time. But getting here taught me things that no Microsoft whitepaper or partner presentation ever mentioned. Here's what I'd tell myself if I could go back to day one.

The Promises vs. The Reality

Let me start with the vendor promises that actually held up. D365 F&O's catch weight functionality is legitimately good for food distribution. When you're selling beef by the case but pricing by the pound, this matters. The integration with Power Platform is real and useful. And the Azure backbone means we're not worrying about infrastructure anymore.

Now the promises that didn't quite land. "Out of the box" is a myth. Every food distribution company has unique processes, and you're going to customize more than the sales deck suggests. The "modern UI" is modern compared to S2K from 2005, but your warehouse team isn't going to mistake it for Instagram. And the implementation timeline your partner quoted? Add 40%.

Data Migration is Where Projects Die

Our legacy system was VAI S2K. Twenty years of data. Customer records with addresses that no longer exist. Pricing agreements from salespeople who retired a decade ago. Items that haven't sold since 2018 but somehow still have inventory in the system.

Here's what I wish someone had told me: clean your data before you start, not during. We spent three months in the middle of the project doing data cleanup that should have happened before we signed the contract. Every week of delay cost money and patience.

The specific pain points:

• Customer hierarchies that made sense in S2K but don't map to D365's structure
• Historical pricing data that conflicts with current agreements
• Item attributes that were free-text in the old system but need to be structured in the new one
• Vendor records with duplicate entries that the business never cleaned up because "it worked"

If I could do it again, I'd spend three months on data cleanup before touching D365. Create the mappings. Identify the conflicts. Make the business decisions about what to keep, what to archive, what to delete. Then migrate.

Integration Architecture Decisions Haunt You

We have BFC Dakota for WMS, Adobe Commerce for B2B e-commerce, Shopify Plus for B2C, and Paycom for HR/payroll. Every one of those integrations was more complex than expected.

The D365 dual-write feature is powerful but temperamental. When it works, it's magic. Real-time sync between Dataverse and F&O. When it doesn't work, you're debugging at 2 AM wondering why customer records aren't flowing.

Our Adobe Commerce integration required custom middleware. The standard connectors assume you're doing things the standard way. We're not. We have customer-specific pricing, order guides, catch-weight products, and complex delivery routing. None of that fits neatly into pre-built integrations.

What I'd do differently: map every integration touchpoint before writing a single line of code. Document the data that needs to flow, the timing requirements, the error handling scenarios. Then build. We did it the other way around and paid for it in rework.

Change Management is Not Optional

Our warehouse team has been using RF scanners with S2K for 15 years. They can pick orders in their sleep. Asking them to learn a new system while maintaining the same productivity targets was asking for trouble.

We underinvested in training. I'll own that. We did the standard "train the trainer" approach and assumed knowledge would cascade. It didn't. Two months before go-live, we discovered that half the warehouse didn't know how to handle exceptions in the new system. We scrambled.

The lesson: train like your go-live depends on it, because it does. Hands-on practice, not PowerPoints. Real scenarios, not sanitized demos. And give people time to make mistakes in a sandbox before it counts.

The Catch Weight Reality

For anyone in food distribution evaluating D365, here's the honest truth about catch weight. It works. But it requires discipline.

Your items need to be set up correctly from the start. The nominal weight, the variance tolerances, the pricing unit versus the inventory unit. Get any of that wrong and you'll have invoicing nightmares.

We had to retrain our receiving team on how to capture actual weights versus nominal weights. The old system was more forgiving of sloppy data entry. D365 is not. That's actually a good thing long-term, but it hurt in the transition.

The pricing engine handles catch weight well once configured. Variable weight products priced by the pound, sold by the case, invoiced accurately. It's genuinely impressive when it's working. The setup to get there is what nobody warns you about.

Power Platform is the Secret Weapon

The best decision we made was investing in Power Platform alongside F&O. Power Apps for custom screens our users actually wanted. Power Automate for approval workflows that used to require emails and signatures. Power BI for dashboards that update in near real-time.

Our sales team now has a mobile app for checking inventory and customer history that they built themselves with my guidance. Our AP team has an invoice approval workflow that eliminated the paper routing slips. Our executives have dashboards they can actually understand without calling IT.

This is where Microsoft's strategy pays off. F&O is the backbone. Power Platform is the nervous system. Together, they're more than either one alone.

What Nobody Tells You About Support

Microsoft Premier Support is fine for documented issues. For the weird edge cases you'll inevitably hit, you need a partner who's seen them before. Our implementation partner has been good, but the real value came from their consultants who've done food distribution implementations specifically.

Also, budget for post-go-live support. You'll need it. We planned for two months of hypercare. We're on month four and still finding things. Some are bugs. Most are processes that weren't fully thought through. Having support resources available to adjust configurations and add small customizations has been essential.

The 10% Remaining

What's left? Final data validation, user acceptance testing round two, and parallel running with the old system. We'll cut over in Q1. The warehouse will grumble for a month, then they'll wonder how they ever lived without real-time inventory visibility.

Was it worth it? Ask me in six months. Right now, I'm confident we'll get value from this investment. Better data, better integrations, better scalability. But the journey was harder than anyone admitted at the start.

Advice for the Next CIO

If you're starting a D365 F&O implementation, here's my checklist:

1. Clean your data first. Really clean it.
2. Map every integration before you build anything.
3. Budget 40% more time than your partner quotes.
4. Invest in change management like it matters, because it does.
5. Learn Power Platform. It's the multiplier.
6. Plan for post-go-live support. The project doesn't end at go-live.
7. Find a partner who knows your industry, not just the software.

D365 F&O is a solid platform. The move from S2K was necessary and will pay dividends. But the implementation is a project, not a product installation. Treat it accordingly.

Now if you'll excuse me, I have a parallel testing cycle to review. The 90% is the hard part. The last 10% is just exhausting.